Testing

The ADK uses standard xUnit for unit and end-to-end testing. The testing suite focuses on verifying the execution of the agent pipeline, tool invocation reliability, processor order operations, and integration via the A2A protocol.

Running Tests

To run the entire suite:

cd Adk/GoogleAdk
dotnet test

Structure

• GoogleAdk.Core.Tests: Contains unit tests for individual processors, memory services, telemetry tracking, and core logic primitives.
• GoogleAdk.E2e.Tests: Contains tests that simulate full multi-turn conversational flows, A2A REST HTTP endpoints, and complete orchestration pipelines like Sequential and Parallel agents.

Testing with InMemory Services

Because the ADK abstracts persistent storage into SessionService, ArtifactService, and MemoryService, writing tests for your complex agent orchestrations is extremely easy without requiring mock databases.

You can rapidly test your agent behavior using the InMemoryRunner.

[Fact]
public async Task CustomAgent_ExecutesTool_AndReturnsSummary()
{
    var agent = new LlmAgent(new LlmAgentConfig { ... });

    // Bootstraps isolated in-memory stores perfect for testing 
    var runner = new InMemoryRunner("test-app", agent);
    var session = await runner.SessionService.CreateSessionAsync(new CreateSessionRequest { AppName = "test-app", UserId = "user-1" });

    var userMessage = new Content { Role = "user", Parts = [new Part { Text = "Test query" }] };

    var responseReceived = false;
    await foreach (var evt in runner.RunAsync("user-1", session.Id, userMessage))
    {
        if (evt.IsFinalResponse() && evt.Content?.Parts != null)
        {
            responseReceived = true;
        }
    }

    Assert.True(responseReceived);
}